Artificial Intelligence-Based Geological Exploration Methods and Systems

Abstract

This invention relates to the field of geological exploration technology, and discloses a geological exploration method and system based on artificial intelligence. The method includes: acquiring geological data of a target area and performing feature decoupling to obtain independent feature components of the target area; performing regular mapping on the correlation between independent feature components to obtain coupling relationships; constructing a feature association graph of the target area using independent feature components as nodes and the coupling relationships as edges; performing neighborhood semantic aggregation on the feature association graph to obtain fused node features; performing logical induction on the fused node features to obtain geological differentiation patterns, and performing zoning deduction to obtain a geological division scheme for the target area; and evaluating the contribution of the feature association graph to obtain the main controlling factors of the geological structure of the target area. This invention can improve the efficiency of geological exploration based on artificial intelligence.

Description

Technical Field

[0001] This invention relates to the field of geological exploration technology, and in particular to geological exploration methods and systems based on artificial intelligence. Background Technology

[0002] Current geological exploration technologies have significant shortcomings in data preprocessing and feature extraction. They lack efficient noise filtering mechanisms for collected geological datasets, and standardization is insufficient. They cannot effectively separate feature components from geological data using precise blind source separation techniques, and the independence verification process is incomplete, making it difficult to obtain independent feature components for the target region. Furthermore, there is a lack of scientific coupling quantification methods for the correlation between independent feature components, and structured association rules are missing, making it impossible to achieve an effective nonlinear mapping transformation from initial correlation strength to coupling strength. This results in an ambiguous representation of the coupling relationships between independent feature components.

[0003] Current technologies lack a systematic framework for constructing feature association graphs, and the weight calculation of feature nodes lacks a reasonable basis. Furthermore, the semantic aggregation process fails to fully integrate the importance differences between node attributes and neighborhood features, resulting in fused node features that cannot comprehensively reflect the actual geological situation. In addition, the summarization of geological differentiation patterns lacks a rigorous logical process, the optimization of candidate partitions and boundary fitting during zoning deduction are not precise enough, and the identification of controlling factors lacks scientific support based on feature subgraph structure analysis and contribution quantification. This leads to poor adaptability of geological division schemes to the actual geological structure of the target region, resulting in insufficient reliability and practicality of the survey results. Therefore, improving the efficiency of AI-based geological surveys has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a geological survey method and system based on artificial intelligence to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides an artificial intelligence-based geological survey method, comprising: S1. Obtain geological data of the target region and perform feature decoupling on the geological data to obtain independent feature components of the target region; S2. Perform regularized mapping on the correlation between the independent feature components to obtain the coupling relationship between the independent feature components; S3. Construct a feature association graph of the target region using the independent feature components as nodes and the coupling relationships as edges; S4. Perform neighborhood semantic aggregation on the feature association graph to obtain the fusion node features of the target region; S5. Logically summarize the features of the fusion nodes to obtain the geological differentiation pattern of the fusion node features, and perform regionalization deduction on the geological differentiation pattern to obtain the geological division scheme of the target region. S6. Based on the geological classification scheme, evaluate the contribution of the feature association map to obtain the main controlling factors of the geological structure of the target region.

[0006] In a preferred embodiment, acquiring geological data of the target region and decoupling the geological data to obtain independent feature components of the target region includes: Collect geological datasets for the target region; The geological dataset is subjected to noise filtering to obtain standardized geological data for the target region; The standardized geological data is subjected to blind source separation to obtain the feature components of the geological data; The independence of the feature components is verified to obtain the independent feature components of the target region.

[0007] In a preferred embodiment, the step of regularizing the mapping of the correlation relationships between the independent feature components to obtain the coupling relationships of the independent feature components includes: The independent feature components are coupled and quantized to obtain the correlation strength of the independent feature components; Statistical analysis of the association strength is performed to obtain the structured association rules for the association strength; Based on the structured association rules, the initial association strength is nonlinearly mapped and transformed to obtain the coupling strength of the independent feature components; Correlation analysis is performed on the coupling strength to obtain the coupling relationship of the independent feature components.

[0008] In a preferred embodiment, constructing a feature association graph of the target region using the independent feature components as nodes and the coupling relationships as edges includes: Geological attribute encoding is performed on the independent feature components to obtain the feature nodes of the independent feature components; Calculate the weight coefficients of the feature nodes; ; in, For the feature node The weighting coefficients, The correlation strength value between the feature nodes. This represents the sum of association strengths in the feature association graph. The traversal index for all nodes in the feature association graph; The weighted coefficients are weighted to the correlation strength of the coupling relationship to obtain the weighted edges of the feature nodes; The feature nodes and the weighted edges are structurally assembled to obtain a preliminary feature association map of the target region; Spatial continuity optimization is performed on the preliminary feature association map to obtain the feature association map of the target region.

[0009] In a preferred embodiment, the step of structurally assembling the feature nodes and the weighted edges to obtain a preliminary feature association graph of the target region includes: The association relationship is established for the node pairs connected by the weighted edges to obtain the topological skeleton of the feature nodes; Based on the topological skeleton, structural analysis is performed on the feature nodes and the weighted edges to obtain the topological structure of the target region; The isolated units of the topology are eliminated to obtain a preliminary feature association map of the target region.

[0010] In a preferred embodiment, the step of performing neighborhood semantic aggregation on the feature association graph to obtain the fused node features of the target region includes: The neighborhood traversal of the feature nodes in the feature association graph is performed to obtain the neighborhood features of the feature nodes; Based on the correlation strength of the weighted edges, the neighborhood features are weighted according to their importance to obtain the weighted neighborhood of the neighborhood features; The weighted neighborhood and the attributes of the feature nodes are fused to obtain the fused node features of the target region. The calculation formula for the fused node features is as follows: ; In the formula, For the feature node The fusion node feature values, For dynamic fusion coefficients, For the feature node Its own attribute value, For the feature node The weighting coefficients, For the feature node and The coupling strength, The value is the neighborhood feature value of the feature.

[0011] In a preferred embodiment, the step of logically summarizing the features of the fused nodes to obtain the geological differentiation patterns of the fused node features, and then performing regionalization deduction on the geological differentiation patterns to obtain a geological division scheme for the target region, includes: Clustering is performed on the features of the fused nodes to obtain clusters of similar attribute features of the fused nodes; By performing regularity deduction on the similar attribute feature clusters, the geological differentiation patterns of the similar attribute feature clusters are obtained; A comprehensive analysis of the geological differentiation patterns yields a geological classification scheme for the target region.

[0012] In a preferred embodiment, the step of comprehensively analyzing the geological differentiation patterns to obtain a geological classification scheme for the target region includes: Based on the geological differentiation pattern, spatial boundaries are extracted from the feature association map to obtain candidate partitions of the target region; The candidate partitions are optimized for topological continuity to obtain the optimized partitions for the target region. Boundary fitting is performed on the optimized partition to obtain the geological partition boundaries of the target region; By comprehensively planning the geological zoning boundaries and the feature association map, a preliminary division scheme for the target region is obtained; The preliminary division scheme is adapted and optimized to obtain the geological division scheme of the target region.

[0013] In a preferred embodiment, the step of evaluating the contribution of the feature correlation map based on the geological classification scheme to obtain the main controlling factors of the geological structure of the target region includes: Based on the geological division scheme, the feature association map is sub-mapped to obtain the feature sub-map of the target region; Structural analysis is performed on the feature subgraph to obtain its internal connection structure and intensity distribution; Based on the internal connection structure and strength distribution, the contribution of the independent feature components is quantified to obtain the contribution degree of the independent feature components; The dominant factors of the contribution are identified to obtain the main controlling factors of the geological structure of the target region.

[0014] To address the above problems, the present invention also provides an artificial intelligence-based geological exploration system, the system comprising: The data feature decoupling module is used to acquire geological data of the target region and decouple the geological data to obtain independent feature components of the target region. The feature association rule mapping module is used to perform rule-based mapping on the association relationships between the independent feature components to obtain the coupling relationships of the independent feature components; The feature association graph construction module is used to construct the feature association graph of the target region using the independent feature components as nodes and the coupling relationships as edges. The node neighborhood aggregation module is used to perform neighborhood semantic aggregation on the feature association graph to obtain the fused node features of the target region; The geological zoning deduction module is used to logically summarize the characteristics of the fusion nodes, obtain the geological differentiation rules of the fusion node characteristics, and perform zoning deduction on the geological differentiation rules to obtain the geological division scheme of the target region. The contribution assessment module is used to assess the contribution of the feature association map based on the geological classification scheme, so as to obtain the main controlling factors of the geological structure of the target region.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention employs a series of refined processes, including feature decoupling and regular mapping of correlation relationships, to accurately capture independent geological feature components and their coupling relationships. The construction of the feature correlation graph combines weight coefficient calculation with spatial continuity optimization, resulting in a more comprehensive and realistic representation of geological feature correlations. The integration of dynamic fusion coefficients, weight coefficients, and other multi-dimensional parameters during neighborhood semantic aggregation achieves efficient fusion of feature node attributes with neighborhood features, significantly improving the accuracy of fused node features and providing high-quality data support for geological pattern mining.

[0016] 2. By relying on the logical induction and zoning deduction of fused node features, this invention can accurately extract geological differentiation patterns. Through steps such as candidate partition optimization and boundary fitting, it generates a scientifically sound geological division scheme, making the geological partitions more closely aligned with the actual geological conditions of the region. Simultaneously, based on the contribution assessment of the geological division scheme, through feature subgraph structure analysis and contribution quantification, it can accurately identify the main controlling factors of geological structure, significantly improving the reliability and relevance of geological survey results, providing strong support for geological-related decision-making, and overall enhancing the scientific rigor and efficiency of geological surveys. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating an artificial intelligence-based geological survey method according to an embodiment of the present invention. Figure 2 A functional block diagram of an artificial intelligence-based geological survey system provided in an embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides an artificial intelligence-based geological survey method. The executing entity of the artificial intelligence-based geological survey method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application embodiment: a server, a terminal, etc. In other words, the artificial intelligence-based geological survey method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating an artificial intelligence-based geological survey method according to an embodiment of the present invention. In this embodiment, the artificial intelligence-based geological survey method includes: S1. Obtain geological data of the target region and perform feature decoupling on the geological data to obtain independent feature components of the target region; In this embodiment of the invention, the step of acquiring geological data of the target region and performing feature decoupling on the geological data to obtain independent feature components of the target region includes: Collect geological datasets for the target region; The geological dataset is subjected to noise filtering to obtain standardized geological data for the target region; The standardized geological data is subjected to blind source separation to obtain the feature components of the geological data; The independence of the feature components is verified to obtain the independent feature components of the target region.

[0021] When collecting geological datasets for a target region, the specific scope of the target region is first defined. Boundaries are delineated using latitude and longitude, and the covered geographical area is determined, including different topographic units such as mountains, plains, and basins. Then, the types of geological data to be collected are determined, covering core data related to geological characteristics, such as stratigraphic and lithological data, geological structural data, geophysical exploration data, and geochemical analysis data. A combination of field surveys and existing data integration is employed. Field surveys involve obtaining core samples at different depths through geological drilling, recording information such as lithology, structure, and water-bearing capacity of the cores. Outcrop observations record data such as the occurrence of exposed strata, rock types, and geological structural traces. Simultaneously, rock, soil, and groundwater samples are collected for subsequent analysis. Existing data integration involves extracting historical geological data relevant to the target region from official databases of geological survey departments, geological research reports published by research institutions, and archived survey data within the industry, ensuring that the data covers the main geological units and key geological features of the target region. Finally, all the first-hand data collected in the field and the integrated historical data were classified and sorted out according to data type, collection location, collection time and other dimensions to form a complete geological dataset of the target region.

[0022] When filtering noise from geological datasets, a comprehensive analysis of potential noise types is first conducted. This includes error data caused by insufficient instrument precision, deviation data resulting from extreme weather or terrain interference during field collection, and erroneous data caused by human input errors during data recording. For instrument error noise, data from multiple measurements at the same collection point using the same instrument are compared, retaining values ​​with reasonable fluctuations and removing abnormal data that significantly deviates from the majority of measurement results. For environmental interference noise, environmental records from the collection process, such as rainfall, strong winds, and terrain obstruction, are considered to remove geological data that is abnormal due to these factors. For human error noise, the original data recording documents and collection logs are cross-checked, and erroneous data is corrected; invalid data whose authenticity cannot be verified is directly removed. After noise filtering, all valid data is converted to a unified unit system, such as standardizing length to meters, mass to grams, and concentration to milligrams per liter. Simultaneously, the data is standardized and organized according to fixed field formats to ensure complete consistency in the recording methods and field definitions for each type of geological data, ultimately yielding standardized geological data for the target region.

[0023] When performing blind source separation on standardized geological data, it is first recognized that the standardized geological data is a mixture of geological information dominated by multiple different geological factors. This mixture consists of superimposed original geological signals. The core of blind source separation is to decompose the mixed signal into independent original signals without knowing the specific attributes of each original geological signal. These original signals are the characteristic components of the geological data. In practice, the standardized geological data is first divided into multiple data sequences according to the spatial distribution or temporal variation of geological phenomena. Each data sequence reflects the changes in geological information within a specific spatial range or time period. Then, the variation patterns of each data sequence are analyzed. By comparing the numerical fluctuation trends, peak frequency, and numerical variation amplitude of different data sequences, signal components with unique variation patterns are identified. These signal components correspond to different geological influencing factors, such as signal components dominated by stratigraphic lithology differences, signal components dominated by geological tectonic activity, and signal components dominated by geochemical element distribution. By gradually separating these signal components with independent variation patterns and eliminating signal interference between different geological factors, multiple independent characteristic components of the geological data are finally obtained.

[0024] When performing independence checks on feature components, multiple check dimensions are determined, including the data variation trend, numerical distribution characteristics, and correlation with geological genesis. First, the data patterns of each feature component are analyzed to check whether its changes in spatial or temporal dimensions are consistent and whether they conform to the mechanism of action of the corresponding geological factors. For example, feature components reflecting stratigraphic lithology should have numerical variations consistent with the vertical distribution or horizontal extension patterns of the strata. Then, the independence between different feature components is determined by comparing the numerical distribution curves and variation trend curves of each feature component to observe whether there are overlapping areas or obvious dependencies. If the variation trends of two feature components are highly consistent and the numerical distributions show a clear correlation, it indicates a dependency, and the process needs to return to the blind source separation step for reprocessing. If the variation trends of each feature component are unrelated, the numerical distributions do not overlap, and each feature component clearly corresponds to a unique geological genesis without mutual interference, then the feature component is determined to be independent. All feature components that pass the independence check are summarized and organized to obtain the independent feature components of the target region.

[0025] The beneficial effects are that the above implementation process ensures the comprehensiveness and integrity of the dataset through standardized geological data collection, precise noise filtering removes various interference factors, making the standardized geological data highly accurate and reliable, successful blind source separation decomposes mixed geological information into feature components of a single cause, and independence verification further ensures the purity and independence of each feature component. The resulting independent feature components can accurately and comprehensively reflect the essential characteristics of different geological factors in the target area, providing high-quality data support for subsequent core geological exploration work such as independent feature component correlation analysis and feature correlation map construction, effectively improving the scientific nature of the geological exploration process and the accuracy of subsequent exploration results.

[0026] S2. Perform regularized mapping on the correlation between the independent feature components to obtain the coupling relationship between the independent feature components; In this embodiment of the invention, the step of regularizing the mapping of the correlation between the independent feature components to obtain the coupling relationship of the independent feature components includes: The independent feature components are coupled and quantized to obtain the correlation strength of the independent feature components; Statistical analysis of the association strength is performed to obtain the structured association rules for the association strength; Based on the structured association rules, the initial association strength is nonlinearly mapped and transformed to obtain the coupling strength of the independent feature components; Correlation analysis is performed on the coupling strength to obtain the coupling relationship of the independent feature components.

[0027] Collect specific attribute data of all independent feature components within the target region, including the specific values ​​and manifestations of each independent feature component at different sampling locations. For each pair of independent feature components, analyze the correlation of their numerical changes at all sampling locations. By observing the degree to which the value of one feature component changes when the value of the other feature component changes, convert this degree of correlation into a specific quantitative result. This quantitative result is the correlation strength of the independent feature components.

[0028] The correlation strengths between all independent feature components are arranged in order of numerical value. Then, the distribution of correlation strengths within different numerical ranges is counted. At the same time, the attribute types of the two independent feature components corresponding to each correlation strength are recorded. The common characteristics of feature components with the same attribute type combination on the correlation strength value are analyzed, as well as the differences in the influence of feature components with different attribute type combinations on the correlation strength value. Based on these statistical analysis results, clear and regular clauses are formulated. These clauses clarify the correlation strength range and expression form corresponding to feature components with different attribute combinations, which are the structured association rules of correlation strength.

[0029] Referring to the requirements for adjusting the association strength of feature components with different attribute combinations in the structured association rules, and combining the actual characteristics of the interaction between independent feature components, the initial association strength is numerically transformed. During the transformation process, the initial association strength of different attribute combinations and different numerical ranges is adjusted differently according to the proportions and standards specified in the rules, so as to fully reflect the nonlinear characteristics of the interaction between feature components. After this targeted transformation, the obtained value is the coupling strength of the independent feature components.

[0030] The coupling strength values ​​between all independent feature components are comprehensively analyzed, and the interaction state reflected by the coupling strength of each pair of independent feature components is analyzed one by one. Based on the specific values ​​of the coupling strength and the corresponding feature component attributes, it is determined whether each pair of feature components is mutually promoting, mutually inhibiting, or synergistic. All judgment results are systematically sorted out to clarify the specific interaction mode between each pair of independent feature components, thereby obtaining the coupling relationship of independent feature components.

[0031] The beneficial effects are that by systematically coupling and quantifying, statistically analyzing, nonlinearly mapping and transforming, and analyzing the correlation of independent feature components, the correlation strength, structured correlation rules, coupling strength, and coupling relationship of independent feature components are accurately obtained. This solves the problem of fuzzy representation of the coupling relationship of independent feature components in existing technologies, provides accurate and reliable core data for the construction of subsequent feature correlation maps, and ensures that the feature correlation maps can truly reflect the intrinsic connections between geological features of the target area. This improves the scientificity and accuracy of the entire geological exploration process and lays a solid foundation for the subsequent summarization of geological differentiation laws and the formulation of geological classification schemes.

[0032] S3. Construct a feature association graph of the target region using the independent feature components as nodes and the coupling relationships as edges; In this embodiment of the invention, constructing a feature association graph of the target region using the independent feature components as nodes and the coupling relationships as edges includes: Geological attribute encoding is performed on the independent feature components to obtain the feature nodes of the independent feature components; Calculate the weight coefficients of the feature nodes; ; in, For the feature node The weighting coefficients, The correlation strength value between the feature nodes. This represents the sum of association strengths in the feature association graph. The traversal index for all nodes in the feature association graph; The weighted coefficients are weighted to the correlation strength of the coupling relationship to obtain the weighted edges of the feature nodes; The feature nodes and the weighted edges are structurally assembled to obtain a preliminary feature association map of the target region; Spatial continuity optimization is performed on the preliminary feature association map to obtain the feature association map of the target region.

[0033] The step of structurally assembling the feature nodes and the weighted edges to obtain a preliminary feature association graph of the target region includes: The association relationship is established for the node pairs connected by the weighted edges to obtain the topological skeleton of the feature nodes; Based on the topological skeleton, structural analysis is performed on the feature nodes and the weighted edges to obtain the topological structure of the target region; The isolated units of the topology are eliminated to obtain a preliminary feature association map of the target region.

[0034] When encoding the geological attributes of independent feature components, the specific geological attribute type corresponding to each independent feature component is first determined, including stratigraphic lithology attributes, geological structural attributes, geophysical response attributes, geochemical composition attributes, etc. Then, core descriptive information is extracted for each geological attribute, such as rock type, grain size, and genetic type of stratigraphy, and fault type, fold morphology, and development degree of geological structures. Subsequently, this geological attribute information is converted into feature identifiers in a unified format. Each identifier uniquely corresponds to an independent feature component, and the identifier contains the key information of the geological attribute of that component. In this way, the geological attribute encoding of all independent feature components is completed, and the feature nodes of the independent feature components are obtained.

[0035] The weight coefficients of the feature nodes are calculated to determine their relative importance within the entire feature association system. This determines the weight coefficient for each feature node, ensuring that the weight coefficients accurately reflect the influence of the feature node on the geological association. The association strength value is obtained by coupling and quantizing independent feature components. The sum of association strengths is calculated by sequentially adding the association strength values ​​between all nodes in the feature association graph. The traversal index covers each node in the feature association graph one by one.

[0036] When calculating the weight coefficient of a feature node, the established correlation strength between each feature node and all other feature nodes is first reviewed. The correlation strength of each feature node with each other node is then summed one by one to obtain the total correlation strength of the feature node. Then, the sum of all correlation strengths between all feature nodes in the target region is calculated. By comparing the total correlation strength of a single feature node with the sum of the correlation strengths of all nodes, the relative importance of the feature node in the entire feature association system is determined. This determines the weight coefficient corresponding to each feature node, ensuring that the weight coefficient can truly reflect the influence of the feature node on the geological association relationship.

[0037] The higher the feature node's own attribute value, the higher its weight coefficient, the stronger its coupling with neighboring nodes, and the higher the neighborhood feature value, and the more appropriate the dynamic fusion coefficient is set, the higher the fused node feature value will be. Conversely, the lower the above data, the lower the fused node feature value will be. At the same time, when the dynamic fusion coefficient increases, the influence of the feature node's own attributes on the fused node feature result will increase. When the dynamic fusion coefficient decreases, the influence of the weighted neighborhood features on the fused node feature result will increase.

[0038] When weighting the coefficients to the correlation strength of the coupling relationship, first determine the weight coefficient corresponding to each feature node and the correlation strength of the coupling relationship between the feature node and other feature nodes. Then, combine the weight coefficients of the two feature nodes involved in each coupling relationship with the correlation strength of the coupling relationship. This makes the correlation strength not only reflect the degree of correlation between the two nodes, but also incorporate the importance of the node itself. After such combination processing, the original correlation strength of the coupling relationship is transformed into a new strength value with the influence of node weights, thereby obtaining the weighted edge of the feature node.

[0039] When establishing associations between node pairs connected by weighted edges, all weighted edges are first sorted out, and the feature nodes connected at both ends of each weighted edge are identified, forming several node pairs. Then, for each pair of nodes, the association nature established by them through the weighted edge is clarified, such as causal association, adjoint association, or constraint association. At the same time, the strength information of each weighted edge is recorded. The association nature of all node pairs and the corresponding weighted edge strengths are organized according to geological logic to construct a basic framework that can reflect the core connection relationship between feature nodes, thus obtaining the topological skeleton of the feature nodes.

[0040] Based on the topological skeleton, when performing structural analysis on the feature nodes and weighted edges, the topological skeleton is used as the core framework. The position of each feature node in the skeleton, the number of weighted edges connected to the node, and the strength level of each weighted edge are analyzed one by one. At the same time, the hierarchical relationship between feature nodes formed by weighted edges is clarified, such as which nodes are in the core connection position and which nodes are in the secondary connection position. Then, the position information, connection relationship, weighted edge strength information, and hierarchical relationship of all feature nodes are comprehensively integrated to sort out the overall organizational form formed by the cooperation of feature nodes and weighted edges, and obtain the topological structure of the target region.

[0041] When eliminating isolated units in the topology, all feature nodes in the topology are first thoroughly investigated to identify individual feature nodes that are not associated with other feature nodes through any weighted edges, as well as sets of nodes that are connected to each other but not associated with any weighted edges of the main body of the topology. These are all isolated units. Subsequently, it is confirmed that these isolated units are not necessary components of the geological features of the target region, but invalid units generated during data processing. Then, these isolated units are completely removed from the topology, retaining the feature nodes and weighted edges that are effectively associated with the main body of the topology, to obtain the preliminary feature association map of the target region.

[0042] When optimizing the spatial continuity of the preliminary feature association map, the actual spatial geographic distribution information of the target region is considered, including the continuous distribution of topography and landforms, the spatial extension pattern of strata and rock layers, etc., to check whether there are any broken connections between feature nodes or situations that do not conform to the spatial distribution logic in the preliminary feature association map. For example, feature nodes within the same geological unit lack necessary weighted edge connections, or there are unreasonable cross-boundary connections between different geological units. To address these issues, weighted edges that conform to spatial logic are added to connect the broken connections, and unreasonable cross-boundary weighted edges are deleted, so that the connections between feature nodes and weighted edges in the map can be consistent with the actual geological spatial distribution of the target region, ensuring the spatial continuity of the association relationship, and thus obtaining the feature association map of the target region.

[0043] The beneficial effects are as follows: geological attribute encoding enables each feature node to accurately correspond to the geological essence of its independent feature components; weighted edges incorporate node weight coefficients to improve the specificity of the association relationship; the construction of the topological skeleton and topological structure clearly presents the connection logic of the feature nodes; isolated unit elimination removes invalid interference factors; spatial continuity optimization makes the association graph conform to the actual geological spatial distribution of the target region; and the final feature association graph can comprehensively, accurately and realistically reflect the association relationship between independent feature components of the target region, providing a high-quality structural foundation for subsequent geological exploration links such as neighborhood semantic aggregation and geological differentiation pattern mining, effectively improving the scientific nature of the entire geological exploration process and the accuracy of the final results.

[0044] S4. Perform neighborhood semantic aggregation on the feature association graph to obtain the fusion node features of the target region; In this embodiment of the invention, the step of performing neighborhood semantic aggregation on the feature association graph to obtain the fused node features of the target region includes: The neighborhood traversal of the feature nodes in the feature association graph is performed to obtain the neighborhood features of the feature nodes; Based on the correlation strength of the weighted edges, the neighborhood features are weighted according to their importance to obtain the weighted neighborhood of the neighborhood features; The weighted neighborhood and the attributes of the feature nodes are fused to obtain the fused node features of the target region. The calculation formula for the fused node features is as follows: ; In the formula, For the feature node The fusion node feature values, For dynamic fusion coefficients, For the feature node Its own attribute value, For the feature node The weighting coefficients, For the feature node and The coupling strength, The value is the neighborhood feature value of the feature.

[0045] For each feature node in the feature association graph, we examine all other feature nodes in the graph that are directly connected to it through weighted edges. These directly connected nodes are identified as the neighboring nodes of the feature node. Then, we collect all the geological attribute information contained in each neighboring node and integrate the collected neighboring node attribute information into a set. This set is the neighboring feature of the feature node.

[0046] For each feature node's neighborhood features, the association strength corresponding to the weighted edge connecting each neighborhood node to the current feature node is determined. The specific situation of the association strength is used as the judgment basis. The higher the association strength, the greater the influence and importance of the attribute information of the corresponding neighborhood node on the current feature node. According to this correspondence, the attribute information of each neighborhood node in the neighborhood features is assigned a corresponding importance weight. The set of neighborhood features after weight assignment is the weighted neighborhood of the neighborhood features.

[0047] First, the inherent geological attribute information of each feature node is extracted and used as the core basic data for the fusion process. Then, the weighted neighborhood corresponding to the feature node is introduced as supplementary data. During the fusion process, the core position of the feature node's own attribute information is fully preserved, and the proportion of neighborhood information in the fusion result is reasonably allocated according to the importance weight of each part of the information in the weighted neighborhood. By comprehensively and coordinately integrating the core basic data and the weighted supplementary data, comprehensive information that can simultaneously reflect the characteristics of the feature node itself and the influence of neighborhood association is formed. This comprehensive information is the fusion node feature of the target region.

[0048] The core significance of this formula is to achieve a scientific fusion of the feature node's own attributes and the weighted neighborhood. By dynamically allocating the proportion of the two in the fusion result through the fusion coefficient, and by fully combining the weight coefficient and coupling strength of the feature node, it accurately reflects the differences in importance of neighborhood features. Finally, it generates fusion node feature values ​​that can comprehensively and realistically reflect the feature node's own characteristics and the influence of neighborhood associations, providing high-quality data support for the subsequent induction of geological differentiation laws.

[0049] The dynamic fusion coefficient is a specially set value during the feature fusion process, used to coordinate the proportion of the feature node's own attribute value and the weighted neighborhood in the fusion result. The feature node's own attribute value is the inherent geological attribute information of the feature node, which is determined simultaneously when the feature node is obtained by encoding the geological attributes of independent feature components. The weight coefficient of the feature node is calculated using a weight of 5. The neighborhood feature value is obtained by traversing the neighborhood of the feature node in the feature association graph. During the traversal, the attribute information of all directly connected neighboring nodes of the feature node is collected, and this information is integrated to form the neighborhood feature value.

[0050] When the feature node's own attribute value increases, the fusion node's feature value also increases, because its own attribute value directly participates in the fusion calculation and occupies a fixed proportion of weight. When the calculation result corresponding to the weighted neighborhood increases, the fusion node's feature value increases synchronously. The weighted neighborhood is the product of neighborhood features weighted by importance, and its numerical change directly affects the fusion result. When the weight coefficient of a feature node increases, it enhances the influence of the neighborhood feature-related calculation part in the fusion result. With the neighborhood feature value and coupling strength remaining constant, the fusion node's feature value increases with the increase of the weight coefficient. When the coupling strength between a feature node and other nodes increases, it strengthens the contribution of neighborhood features to the fusion result. With the neighborhood feature value and weight coefficient remaining constant, the fusion node's feature value increases with the increase of the coupling strength. When the dynamic fusion coefficient increases, the proportion of the feature node's own attribute value in the fusion result increases. If its own attribute value is greater than the calculation result corresponding to the weighted neighborhood, the fusion node's feature value increases; if its own attribute value is less than the calculation result corresponding to the weighted neighborhood, the fusion node's feature value decreases.

[0051] The beneficial effects are as follows: by orderly neighborhood traversal, the neighborhood features of feature nodes are accurately obtained; by weighting the importance of the weighted edge association strength, the value of neighborhood information is reasonably distinguished; and by organic feature fusion, the node's own attributes are fully combined with the weighted neighborhood. The whole process relies on the data processing logic of artificial intelligence, which effectively avoids the problems of insufficient fusion of node attributes and neighborhood features and unclear distinction of importance. The final fused node features can comprehensively and accurately reflect the internal correlation and characteristics of geological features, providing high-quality and reliable data support for the subsequent summarization of geological differentiation laws and the formulation of geological classification schemes, and significantly improving the scientificity and effectiveness of data processing in geological exploration.

[0052] S5. Logically summarize the features of the fusion nodes to obtain the geological differentiation pattern of the fusion node features, and perform regionalization deduction on the geological differentiation pattern to obtain the geological division scheme of the target region. In this embodiment of the invention, the step of logically summarizing the features of the fused nodes to obtain the geological differentiation patterns of the fused node features, and then performing regionalization deduction on the geological differentiation patterns to obtain a geological division scheme for the target region, includes: Clustering is performed on the features of the fused nodes to obtain clusters of similar attribute features of the fused nodes; By performing regularity deduction on the similar attribute feature clusters, the geological differentiation patterns of the similar attribute feature clusters are obtained; A comprehensive analysis of the geological differentiation patterns yields a geological classification scheme for the target region.

[0053] The comprehensive analysis of the geological differentiation patterns to obtain the geological division scheme for the target region includes: Based on the geological differentiation pattern, spatial boundaries are extracted from the feature association map to obtain candidate partitions of the target region; The candidate partitions are optimized for topological continuity to obtain the optimized partitions for the target region. Boundary fitting is performed on the optimized partition to obtain the geological partition boundaries of the target region; By comprehensively planning the geological zoning boundaries and the feature association map, a preliminary division scheme for the target region is obtained; The preliminary division scheme is adapted and optimized to obtain the geological division scheme of the target region.

[0054] When clustering the features of fusion nodes, we first identify the core geological attribute dimensions covered by the features of fusion nodes, including stratigraphic and lithological assemblage characteristics, geological structural response characteristics, geochemical element enrichment characteristics, and geophysical field anomaly characteristics. We then compare the specific performance of each feature of fusion nodes in these dimensions, such as the combination of rock types in stratigraphic and lithological assemblage, the strength of fault influence in structural response, the types and levels of element enrichment, and the amplitude range of physical field anomalies. We group the features of fusion nodes that are completely consistent in all core attribute dimensions, have highly similar geological significance, and have synchronized data change trends. Each group forms a set with a unified geological attribute representation, and finally, we obtain the similar attribute feature clusters of fusion node features.

[0055] When extrapolating patterns from clusters of similar attribute features, for each cluster, a comprehensive analysis is conducted on the common geological attributes of all fused node features within the cluster. This analysis delves into the spatial distribution coordinates of these features within the target region, their spatial relative positions to other clusters of similar attribute features, the weighted edge correlation strength between nodes within and outside the cluster, and the attribute fluctuation characteristics as geographical parameters such as latitude, longitude, and altitude change. The results summarize the spatial distribution differences, numerical trends, and interactions with the surrounding geological environment of the geological attributes corresponding to this cluster. These universal conclusions, derived through systematic analysis, constitute the geological differentiation patterns of clusters of similar attribute features.

[0056] Based on geological differentiation patterns, when extracting spatial boundaries from feature association maps, each geological differentiation pattern clearly corresponds to the distribution range and change boundary of a specific geological attribute. In the feature association map, all feature nodes that conform to the geological differentiation pattern and their interconnected weighted edges are accurately identified. The spatial coordinate correlation of these feature nodes is traced to determine the spatial range covered by the effective connection of the weighted edges. Combined with the attribute change critical point defined by the geological differentiation pattern, the approximate spatial boundary of the concentrated distribution area of ​​this type of geological attribute is clearly delineated. Each such area is a candidate partition of the target region.

[0057] When optimizing the topological continuity of candidate partitions, the connection between feature nodes and weighted edges within each candidate partition is checked one by one. It is checked whether there are any breaks where feature nodes are not effectively associated with other nodes in the partition through weighted edges, or abnormal connections where the connection logic of weighted edges does not conform to the geological differentiation pattern. For breaks, reasonable weighted edges are added to establish effective connections between nodes based on the geological differentiation pattern corresponding to the candidate partition and the association pattern of surrounding nodes. Abnormal connections are directly eliminated to ensure that all feature nodes in the candidate partition form a continuous, complete topological structure that conforms to geological logic through weighted edges, and finally obtain the optimized partition of the target region.

[0058] When performing boundary fitting on the optimized partition, the precise spatial coordinate data of all feature nodes within the optimized partition are collected, including latitude, longitude, altitude, and other information. The spatial distribution pattern and clustering trend of these coordinate points are analyzed, and a smooth curve that can completely wrap all feature nodes and weighted edges within the optimized partition is drawn. The curve drawing strictly follows the actual spatial distribution density of the nodes. In dense node areas, the curve fits the distribution trajectory of the coordinate points, and in sparse node areas, it is reasonably extended according to the distribution trend of adjacent nodes. This ensures that the curve neither exceeds the feature coverage range of the optimized partition nor misses any key feature nodes, and finally obtains the geological partition boundary of the target region.

[0059] When planning the geological zoning boundaries and feature association maps as a whole, all the obtained geological zoning boundaries are completely superimposed on the feature association map. A comprehensive investigation is conducted to check whether there are any overlaps, intersections, or excessive gaps between the boundaries of each geological zoning. Based on the distribution of the association strength and topological logic of the feature nodes in the feature association map, the positions of the boundaries of each geological zoning are adjusted so that the overall association strength of the feature nodes within each geological zoning is higher than the association strength between zonings. At the same time, it is ensured that all feature nodes are reasonably included in their corresponding geological zoning, without omissions or duplicate assignments, thus forming a preliminary division scheme for the target region.

[0060] When adapting and optimizing the preliminary division scheme, the rationality of each geological zone in the preliminary division scheme is verified one by one by combining the existing geological survey results, outcrop observation data, and borehole core analysis results of the target area. It is checked whether the zone boundaries are consistent with the boundaries of the actual geological units and whether the geological attribute characteristics within the zone are consistent with the actual geological conditions. If there are large deviations between the zone boundaries and the actual geological units, or if the zone contains multiple unrelated geological attribute characteristics, the zone boundaries are finely adjusted or the area is re-divided according to the actual geological data to ensure that the division scheme is fully adapted to the actual geological structure of the target area, and finally the geological division scheme of the target area is obtained.

[0061] The beneficial effects are that the above implementation process achieves scientific classification of integrated node features through precise clustering, accurately extracts geological differentiation patterns through regularity deduction, ensures the integrity and logic of the partitioning through spatial boundary extraction and topological continuity optimization, improves the accuracy of partition boundaries through boundary fitting, and ensures that the partitioning scheme closely fits the actual geological conditions through overall planning and adaptation optimization. The final geological partitioning scheme can truly and accurately reflect the distribution characteristics of geological units in the target area, providing a reliable partitioning basis for subsequent work such as the identification of the main controlling factors of geological structure, effectively improving the practicality and scientificity of geological survey results, and assisting in the accurate formulation of geological-related decisions.

[0062] S6. Based on the geological classification scheme, evaluate the contribution of the feature association map to obtain the main controlling factors of the geological structure of the target region.

[0063] In this embodiment of the invention, the step of evaluating the contribution of the feature association map based on the geological classification scheme to obtain the main controlling factors of the geological structure of the target region includes: Based on the geological division scheme, the feature association map is sub-mapped to obtain the feature sub-map of the target region; Structural analysis is performed on the feature subgraph to obtain its internal connection structure and intensity distribution; Based on the internal connection structure and strength distribution, the contribution of the independent feature components is quantified to obtain the contribution degree of the independent feature components; The dominant factors of the contribution are identified to obtain the main controlling factors of the geological structure of the target region.

[0064] Based on the clearly defined boundaries and geological attribute ranges of each geological zone in the established geological division scheme, the feature nodes corresponding to each zone and the weighted edges between these nodes are accurately located in the feature association map. According to the zone range, all feature nodes and weighted edges belonging to the same zone in the feature association map are completely extracted. Each zone forms an independent sub-map containing all geological feature association information of that zone. These sub-maps are the feature sub-maps of the target region.

[0065] For each feature subgraph, we systematically analyze all the feature nodes and weighted edges between them, clarifying the connection relationships between each feature node and other nodes. We determine whether the nodes are directly connected or indirectly connected through intermediate nodes, thus outlining the overall shape and layout of the node connections in the subgraph. This is the internal connection structure of the feature subgraph. Simultaneously, we record the specific values ​​of the association strength of each weighted edge in the subgraph, statistically analyze the frequency of different association strength values, and clarify the distribution of connections of different strength levels in the subgraph, forming the strength distribution of the internal connections of the feature subgraph.

[0066] By combining the internal connection structure of the feature subgraph, the role of the feature node corresponding to each independent feature component in the connection structure is determined. Feature nodes that are located at the connection hub and can be associated with multiple other nodes have a greater impact on the geological structure. Then, by combining the intensity distribution, the total association intensity of all weighted edges connected to the feature node is summarized. The role level of the node in the structure and the total association intensity are comprehensively calculated according to the set standard and converted into a specific quantitative value. This quantitative value is the contribution of the independent feature component.

[0067] The contribution values ​​of all independent feature components are systematically organized and arranged in descending order. The differences between these values ​​are compared and analyzed to screen out independent feature components whose contribution values ​​are significantly higher than those of others. These feature components play a core and leading role in the formation, development and distribution of the geological structure of the target region, and they are the main controlling factors of the geological structure of the target region.

[0068] The beneficial effects are as follows: by performing precise sub-map segmentation according to the geological division scheme, it is ensured that the feature sub-maps can accurately correspond to specific geological zones. After detailed structural analysis and scientific contribution quantification, the effects of independent feature components are accurately evaluated. Finally, the dominant factor identification accurately locks the main controlling factors of geological structure, which solves the problem that the identification of main controlling factors in existing technologies lacks scientific support based on feature sub-map structural analysis and contribution quantification. This significantly improves the accuracy and reliability of the identification of main controlling factors of geological structure, provides accurate and powerful evidence for the application of geological survey results and related decision-making, and further improves the scientificity and practicality of the artificial intelligence-based geological survey system.

[0069] like Figure 2 The diagram shown is a functional block diagram of an artificial intelligence-based geological survey system provided in an embodiment of the present invention.

[0070] The artificial intelligence-based geological survey system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the artificial intelligence-based geological survey system 100 may include a data feature decoupling module 101, a feature association rule mapping module 102, a feature association graph construction module 103, a node neighborhood aggregation module 104, a geological zoning deduction module 105, and a contribution evaluation module 106. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0071] In this embodiment, the functions of each module / unit are as follows: The data feature decoupling module 101 is used to acquire geological data of the target region and perform feature decoupling on the geological data to obtain independent feature components of the target region. The feature association rule mapping module 102 is used to perform rule-based mapping on the association relationship between the independent feature components to obtain the coupling relationship of the independent feature components; The feature association graph construction module 103 is used to construct the feature association graph of the target region using the independent feature components as nodes and the coupling relationships as edges. The node neighborhood aggregation module 104 is used to perform neighborhood semantic aggregation on the feature association graph to obtain the fused node features of the target region. The geological zoning deduction module 105 is used to logically summarize the characteristics of the fusion node to obtain the geological differentiation law of the fusion node characteristics, and to zonify and deduce the geological differentiation law to obtain the geological division scheme of the target region. The contribution assessment module 106 is used to assess the contribution of the feature association map based on the geological classification scheme, so as to obtain the main controlling factors of the geological structure of the target region.

[0072] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0073] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0074] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0075] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0076] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A geological survey method based on artificial intelligence, characterized in that, The method includes: S1. Obtain geological data of the target region and perform feature decoupling on the geological data to obtain independent feature components of the target region; S2. Perform regularized mapping on the correlation between the independent feature components to obtain the coupling relationship between the independent feature components; S3. Construct a feature association graph of the target region using the independent feature components as nodes and the coupling relationships as edges; S4. Perform neighborhood semantic aggregation on the feature association graph to obtain the fusion node features of the target region; S5. Logically summarize the characteristics of the fused nodes to obtain the geological differentiation patterns of the fused node characteristics, and perform regionalization deduction on the geological differentiation patterns to obtain the geological division scheme of the target region, including: Clustering is performed on the features of the fused nodes to obtain clusters of similar attribute features of the fused nodes; By performing regularity deduction on the similar attribute feature clusters, the geological differentiation patterns of the similar attribute feature clusters are obtained; A comprehensive analysis of the geological differentiation patterns yields a geological classification scheme for the target region. The comprehensive analysis of the geological differentiation patterns to obtain the geological division scheme for the target region includes: Based on the geological differentiation pattern, spatial boundaries are extracted from the feature association map to obtain candidate partitions of the target region; The candidate partitions are optimized for topological continuity to obtain the optimized partitions for the target region. Boundary fitting is performed on the optimized partition to obtain the geological partition boundaries of the target region; By comprehensively planning the geological zoning boundaries and the feature association map, a preliminary division scheme for the target region is obtained; The preliminary division scheme is adapted and optimized to obtain the geological division scheme of the target region; S6. Based on the geological classification scheme, evaluate the contribution of the feature association map to obtain the main controlling factors of the geological structure of the target region.

2. The artificial intelligence-based geological survey method as described in claim 1, characterized in that, The process of acquiring geological data of the target region and decoupling the geological data to obtain independent feature components of the target region includes: Collect geological datasets for the target region; The geological dataset is subjected to noise filtering to obtain standardized geological data for the target region; The standardized geological data is subjected to blind source separation to obtain the feature components of the geological data; The independence of the feature components is verified to obtain the independent feature components of the target region.

3. The artificial intelligence-based geological survey method as described in claim 1, characterized in that, The step of regularizing the mapping of the correlation relationships between the independent feature components to obtain the coupling relationships of the independent feature components includes: The independent feature components are coupled and quantized to obtain the correlation strength of the independent feature components; Statistical analysis of the association strength is performed to obtain the structured association rules for the association strength; Based on the structured association rules, the association strength is transformed by a nonlinear mapping to obtain the coupling strength of the independent feature components; Correlation analysis is performed on the coupling strength to obtain the coupling relationship of the independent feature components.

4. The artificial intelligence-based geological survey method as described in claim 1, characterized in that, The construction of the feature association graph of the target region, using the independent feature components as nodes and the coupling relationships as edges, includes: Geological attribute encoding is performed on the independent feature components to obtain the feature nodes of the independent feature components; Calculate the weight coefficient of the feature node, wherein the formula for calculating the weight coefficient is as follows: ； in, For the feature node The weighting coefficients, The correlation strength value between the feature nodes. This represents the sum of association strengths in the feature association graph. The traversal index for all nodes in the feature association graph; The weighted coefficients are weighted to the correlation strength of the coupling relationship to obtain the weighted edges of the feature nodes; The feature nodes and the weighted edges are structurally assembled to obtain a preliminary feature association map of the target region; Spatial continuity optimization is performed on the preliminary feature association map to obtain the feature association map of the target region.

5. The artificial intelligence-based geological survey method as described in claim 4, characterized in that, The step of structurally assembling the feature nodes and the weighted edges to obtain a preliminary feature association graph of the target region includes: The association relationship is established for the node pairs connected by the weighted edges to obtain the topological skeleton of the feature nodes; Based on the topological skeleton, structural analysis is performed on the feature nodes and the weighted edges to obtain the topological structure of the target region; The isolated units of the topology are eliminated to obtain a preliminary feature association map of the target region.

6. The geological survey method based on artificial intelligence as described in claim 1, characterized in that, The step of performing neighborhood semantic aggregation on the feature association graph to obtain the fused node features of the target region includes: The neighborhood traversal of the feature nodes in the feature association graph is performed to obtain the neighborhood features of the feature nodes; Based on the association strength of the weighted edges corresponding to the feature nodes in the independent feature components, the neighborhood features are weighted according to their importance to obtain the weighted neighborhood of the neighborhood features. The weighted neighborhood and the attributes of the feature nodes are fused to obtain the fused node features of the target region. The calculation formula for the fused node features is as follows: ； In the formula, For the feature node The fusion node feature values, For dynamic fusion coefficients, For the feature node Its own attribute value, For the feature node The weighting coefficients, For the feature node and The coupling strength, The value is the neighborhood feature value of the feature.

7. The artificial intelligence-based geological survey method as described in claim 1, characterized in that, The contribution assessment of the feature correlation map based on the geological classification scheme to obtain the main controlling factors of the geological structure of the target region includes: Based on the geological division scheme, the feature association map is sub-mapped to obtain the feature sub-map of the target region; Structural analysis is performed on the feature subgraph to obtain its internal connection structure and intensity distribution; Based on the internal connection structure and strength distribution, the contribution of the independent feature components is quantified to obtain the contribution degree of the independent feature components; The dominant factors of the contribution are identified to obtain the main controlling factors of the geological structure of the target region.

8. A geological survey system based on artificial intelligence, characterized in that, The system for implementing the artificial intelligence-based geological survey method of claim 1 includes: The data feature decoupling module is used to acquire geological data of the target region and decouple the geological data to obtain independent feature components of the target region. The feature association rule mapping module is used to perform rule-based mapping on the association relationships between the independent feature components to obtain the coupling relationships of the independent feature components; The feature association graph construction module is used to construct the feature association graph of the target region using the independent feature components as nodes and the coupling relationships as edges. The node neighborhood aggregation module is used to perform neighborhood semantic aggregation on the feature association graph to obtain the fused node features of the target region; The geological zoning deduction module is used to logically summarize the characteristics of the fused nodes, obtain the geological differentiation patterns of the fused node characteristics, and perform zoning deduction on the geological differentiation patterns to obtain the geological division scheme of the target region, including: Clustering is performed on the features of the fused nodes to obtain clusters of similar attribute features of the fused nodes; By performing regularity deduction on the similar attribute feature clusters, the geological differentiation patterns of the similar attribute feature clusters are obtained; A comprehensive analysis of the geological differentiation patterns yields a geological classification scheme for the target region. The comprehensive analysis of the geological differentiation patterns to obtain the geological division scheme for the target region includes: Based on the geological differentiation pattern, spatial boundaries are extracted from the feature association map to obtain candidate partitions of the target region; The candidate partitions are optimized for topological continuity to obtain the optimized partitions for the target region. Boundary fitting is performed on the optimized partition to obtain the geological partition boundaries of the target region; By comprehensively planning the geological zoning boundaries and the feature association map, a preliminary division scheme for the target region is obtained; The preliminary division scheme is adapted and optimized to obtain the geological division scheme of the target region; The contribution assessment module is used to assess the contribution of the feature association map based on the geological classification scheme, so as to obtain the main controlling factors of the geological structure of the target region.

Images